🛠️ View Sample SDK on GitHub 💳 Pricing & Signup COMING SOON

A simple YouTube-to-Text API that fetches transcripts from YouTube videos.

• Primary method: Uses youtube_transcript_api with proxy rotation for reliability
• Fallback #1: Uses yt-dlp to scrape auto-captions
• Fallback #2: Downloads audio and uses local Whisper CLI for transcription

The API provides:

• Environment variable loading
• Proxy rotation functionality
• Multiple retry mechanisms
• A Flask API endpoint at /transcript/<video_id> that returns transcription data

Example Input (URL)

https://api.youtubetotext.com/transcript/WTPlp90DvM0

Replace WTPlp90DvM0 with a valid video ID (which you can obtain by opening a Youtube video page and copying the tail of the URL)

Output (JSON)

{
  "source":       "whisper-cpp" | "youtube-transcript-api" | "yt-dlp",
  "language":     "English",
  "language_code":"en",
  "is_generated": true,
  "segments": [
    { "start": 0.00, "duration": 3.24, "text": "Hello and welcome…" },
    …
  ]
}

• source: Source of the transcript (which method succeeded)
• language: Language information
• is_generated: Whether the transcript was auto-generated
• segments: Transcript segments with timing information

Base URL: https://api.youtubetotext.com

---

Quickstart

cURL

curl "https://api.youtubetotext.com/transcript/{VIDEO_ID}?lang={lang}" \
  -H "Authorization: Bearer YOUR_API_KEY"

Python

from youtubetotext.client import YouTubeToText

client = YouTubeToText(api_key="YOUR_API_KEY")
response = client.transcript("VIDEO_ID", lang="es,fr")
for segment in response["segments"]:
    print(segment["text"])

JavaScript (Node.js)

import { getTranscript } from './examples/js_example.js';

(async () => {
  const segments = await getTranscript("VIDEO_ID", "YOUR_API_KEY", "en");
  console.log(segments.map(s => s.text).join("\n"));
})();

---